It has been shown through the study of three kinds of carcinoma that the progeny of malignant stem cells have capacity for differentiation, yielding benign and, in one case, normal progeny. Direction of this naturally occurring differentiation of tumors could serve as an alternative to cytotoxic therapy. To this end, we have focused on blastocyst regulation of embryonal carcinoma. Tumor and colony formation of embryonal carcinoma are controlled by direct cell-cell contact with trophectoderm. Neither blastocoele fluid nor inner cell mass cells regulate colony formation. In addition, the neurula of the 8-1/2-day-old mouse embryo is able to regulate tumor formation of neuroblastoma cells. It is now proposed to extend these studies to seek molecular explanations of the phenomena. Specifically, a search will be made for surface molecules on trophectodermal cells that might induce differentiation of embryonal carcinoma cells through the second messenger system. In the event that close apposition of cells is required to facilitate metabolic cooperation, cybrids will be made to determine if the cytoplasm of the trophectodermal cell is capable of regulating the embryonal carcinoma cell. Similar studies will be done with neuroblastoma cells which appear to be regulated by somite cells. SV40 and polyoma viruses can gain entrance to the nucleus of the embryonal carcinoma cell, but they are incapable of integrating and performing early viral functions. Mutants of polyoma have been isolated that will integrate and function. Molecular explanations for these observations will be sought.